In 1980, researchers with the Coronary Drug Project (CDP) reported a startling finding from their data: participants in the placebo arm of the trial who were relatively adherent to their study medication had a substantially better five-year total mortality experience compared to participants who were less adherent to their placebo pills. Subsequently, similar results were reported in post-hoc analyses from three other cardiovascular clinical trials, with observed reductions in total mortality of 46% to 76%. The implication of these results for health care is profound: for reasons that are poorly understood, adherence to an inactive medication may be associated with marked improvement in health outcomes that exceeds the benefits of most effective medical interventions. This association, however, has not been confirmed in all studies. Investigators with the Lipid Research Clinics Coronary Primary Prevention Trial and the Cardiac Arrhythmia Suppression Trial did not find a similar relationship between adherence to placebo and mortality in their datasets. Why most studies that have examined this issue have found an association while others have not is unknown. In this application, we propose a large secondary-data-analysis study examining placebo groups from several previously reported cardiovascular trials. The data from twelve clinical trials will be obtained from collaborating investigators with the original trials, and through the NHLBI Data Distribution Program; these will be subjected to a uniform analytic strategy to examine the relationship between placebo adherence and several health outcomes. The purpose of these analyses is to examine the consistency of the initial CDP finding, to investigate potential mechanisms among those trials in which this relationship is identified, and to study methodologic and participant differences in the trials that found an association with those that did not. Outcomes to be examined include total mortality, cause-specific mortality, and cause-specific morbidity. The study will conclude with a synthesis of all data to estimate an overall association between placebo adherence and improved health outcomes. The results of this investigation will shed important light on powerful biomedical and non-biomedical determinants of health outcomes that, at present, are insufficiently documented and poorly understood. [unreadable] [unreadable] [unreadable]